Extrapolating from Honeybees to Bumblebees in Pesticide Risk Assessment

Prévia do material em texto

Ecotoxicology 8:3, 147_166 1999.
· 1999 Kluwer Academic Publishers, Boston. Manufactured in the Netherlands.
Extrapolating from Honeybees to Bumblebees in
Pesticide Risk Assessment
HELEN M. THOMPSON*1 AND LYNN V. HUNT2
1National Bee Unit, Central Science Laboratory, Sand Hutton, York YO41 1LZ, United Kingdom
2Department of Zoology, Uni¨ersity of Cambridge, Downing Street, Cambridge CB2 3EJ, United Kingdom
h.tompson@csl.gou.uk
Received 3 December 1998; Accepted 10 January 1999; Revised 26 January 1999
Abstract. Bumblebees are important pollinators of many crops and wild flowers and there are both conservation and economic reasons for taking action to assess the impact of pesticides on bumblebees. Pesticide risk assessments for honeybees are based on hazard ratios which rely on application rates and toxicity data and are unlikely to be appropriate for bumblebees. Bumblebees are active at different times and on different crop species and are, therefore, likely to have different exposure profiles. Unlike honeybees, deaths of bumblebees due to pesticides are unlikely to be reported, since the bees are not kept domestically and will die in small numbers. This paper highlights the differences in the potential risk posed by pesticides to bumblebees from that of honeybees. This is based on their exposure through use of crops and flowering weeds and on available data on toxicity of pesticides. This information is also intended as a source document for information on the foraging behavior and phenology of bumblebees for use in risk assessment for pesticides.
Keywords: Pesticides, risk assessment, bumblebees
Recently, there has been concern about the po-tential impact of pesticides on both long-tongued and short-tongued species of bumblebee
Bombus.. There has been a severe decline in the abundance of bumblebees in the last thirty years, particularly in southern Britain, and it is possible that this is due in part to the use of certain pesticides. Bumblebees are important pollinators of many crops and wild flowers and, therefore, there are both conservation and economic rea-sons for taking action to assess the impact of pesticides on bumblebees.
Pesticide risk assessments for honeybees are based on hazard ratios which rely on application rates and toxicity data and are unlikely to be appropriate for bumblebees. Bumblebees are ac-tive at different times and on different crop
*Address for correspondence: Dr Helen Thompson, NBU, CSL, Sand Hutton, York YO41 1LZ, United Kingdom
species and are, therefore, likely to have different exposure profiles. Unlike honeybees, deaths of bumblebees due to pesticides are unlikely to be reported, since the bees are not kept domestically and will die in small numbers.
This paper highlights the differences in the potential risk posed by pesticides to bumblebees from that of honeybees. This is based on their exposure through use of crops and flowering weeds and on available data on toxicity of pesti-cides. This information is also intended as a source document for information on the foraging be-haviour and phenology of bumblebees for use in risk assessment for pesticides.
1.	The use of crops by bumblebees in the UK
A review of literature was carried out to collate currently available information on the use of crop
148 Thompson and Hunt
plants by bumblebees. All insect-pollinated crops known to be grown in the UK were included and cereal crops were considered as a single group. Table 1 below summarises British insect-pol-linated crop species known to be used by bumble-bees.
Honeybees regularly collect honeydew from aphid-infested cereal crops. No records of bum-blebees collecting honeydew from aphids on cere-als crops were found. However, Bombus lucorum and other species have been reported to collect honeydew, usually on trees, in Russia, Finland, USA and the UK Brian, 1957; Bishop, 1994; Teras, 1985.. Bumblebees have also been ob-served to collect pollen from grasses and other wind-pollinated plants and so they may also do so from cereal crops. Therefore it is likely that bum-blebees use cereal crops to some extent.
Bumblebees were recorded as frequent on most of the crop species listed. There were only nine crop species known to be visited by bumblebees to which they are recorded as occasional visitors. This probably reflects that fact that bumblebees are often at low population densities, can have small colonies and are also likely to be displaced in many of these experiments by domestic honey-bees, with hives kept close to the crop in ques-tion. It seems reasonable to suggest that bumble-bees actually use the majority of our insect-pol-linated crop plants to a large extent taking their often low and very variable population densities into account ..
2.	Use of wild flowers by bumblebees
Bumblebees will also be exposed to pesticides if they are foraging on wild flowers that grow under crops or in field margins. Wild flowers associated with crops are divided into two groups_weeds mostly annuals . which grow in ploughed fields with crops; and field margin species mostly perennials . which grow in unploughed field boundaries and hedgerows. Bumblebees using the former group, arable weeds, are likely to be more exposed to pesticides than bumblebees foraging on field margins, because farmers try to avoid spraying the field margins and hedgerows with pesticides.
2.1.	Arable weeds
The arable weed flora consists of plants growing underneath crops and often up to the field edge. Due to the annual disturbance of ploughing, perennial plants are unable to establish here, with the exception of those which can regenerate from broken root fragments such as field bindweed Con¨ol¨ulus ar¨ensis.. The flora is therefore made up predominantly of annual species, most of which are not favoured by bumblebees. How-ever, any bees foraging on these species will be directly exposed to pesticide sprays.
2.1.1. Arable weed sur¨ey methods. Thirteen pa-pers were found from the last ten years which list common arable weed species Acker and Lutman, 1995; Barr et al., 1996; Clarke et al., 1993; Fir-bank, 1990; Glasgow et al., 1976; Gwynne and Murray, 1985; Knott et al., 1995; Lutman et al., 1995; Marshall, 1989; Mortimer, 1990; Olgivy et al., 1993; Pallutt, 1993; Watkinson and Bo, 1993.. Species lists from each of these papers were com-piled and the most frequently mentioned species were interpreted as the most common, although there are likely to be other species, less com-monly found, that are also used by bumblebees.
All the papers except one Firbank, 1990. dealt primarily with weeds under broadleaved crops. This is because broad-leaved weeds are a much greater proportion of the weed flora than grasses, under broad-leaved crops, due to the use of graminicides. Broadleaf crops surveyed in the pa-pers included sugar beet, oilseed rape, field bean, potatoes, carrots, peas and linseed. Firbank 1990. dealt with weeds under cereal crops and men-tioned no species that were not included in sev-eral of the other papers concerned with the broadleaf weed flora, indicating that there is little difference in broadleaf weed flora under broadleaf and cereal crops.
The results of the literature survey of bumble-bee use of arable weeds are given in Table 2. Weed species are presented in rank order col-umn 2., according to how many of the 13 re-viewed papers they appeared in. These weed species are so common that they are likely to be found under any crop.
Table 1.	Crop species known to be used by Bombus spp.
	
	
	Common
	Flowering
	Visits by
	
	
	Family
	Species
	name
	season
	Bombusb
	References
	
	
	
	
	
	
	
	
	Alliaceae
	Allium ampeloprasuma
	Leek
	July_Aug
	q
	Macgregor, 1976
	
	
	Allium cepaa
	Onion
	June_Sept
	f*
	Free, 1993; Williams and Free, 1974
	
	
	Allium schoenoprasum
	Chives
	June_Aug
	q
	Free, 1993; Fussell and Corbet, 1992
	
	Boraginaceae
	Borago officinalis
	Borage
	May_Sept
	f*
	Osborne, 1994
	
	Compositae
	Helianthus annuus
	Sunflower
	Aug_Oct
	f*
	Free,1993; Banaszak, 1984
	
	
	Lactuca sati¨aa
	Lettuce
	July_Aug
	o
	Free, 1993
	
	Cruciferae
	Brassica napusrcampestris
	Oilseed rape
	Apr_Aug
	f*
	Free, 1993; Williams, 1985; Delbrassine and Rasmont, 1988
	
	
	Brassica oleraceaa
	Cabbage, etc.
	May_Sept
	q
	Free, 1993; Doorn, 1993
	
	
	Raphanus sati¨usa
	Radish
	May_Sept
	f*
	Free, 1993; Patten et al., 1993
	
	
	Sinapis alba
	White mustard
	April_Oct
	q
	Free, 1993
	
	Cucurbitaceae
	Cucurbita sp.
	Pumpkins,
	
	q*
	Free, 1993; Hurd, 1964
	
	
	
	squashes
	
	
	
	
	Ericaceae
	Vaccinium sp.
	Blueberryr
	April_July
	f
	McGregor, 1976; Free, 1993
	
	
	
	bilberry
	
	
	
	
	
	Vaccinium macrocarpon
	Cranberry
	June_Aug
	f*
	Patten et al., 1993; Reader, 1977; McGregor, 1976
	
	Grossulariaceae
	Ribes nigrum
	Blackcurrant
	April_May
	f*
	Free, 1968b; Free, 1993
	
	
	Ribes rubrum
	Redcurrant
	April_May
	o
	Free, 1993; Wilson, 1929
	
	
	Ribes u¨a-crispa
	Gooseberry
	March_May
	f*
	Free, 1993; Wilson, 1929
	
	Labiatae
	La¨andula latifolia
	Lavender
	July_Sept
	o*
	Fussell and Corbet, 1992; Herrera, 1990
	
	
	Ocimum sp a
	Basil
	June_Sept
	f
	Voloshin, 1995
	
	
	Origanum sp.a
	Marjoram
	July_Sept
	q*
	Williams et al., 1993; Fussell and Corbet, 1992
	
	Leguminosae
	Coronilla ¨aria
	Crown vetch
	June_Aug
	o
	Free, 1993
	
	
	Lotus corniculatus
	Birdsfoot
	May_Sept
	f*
	Fussell and Corbet, 1992; Williams, 1997; Patten et al., 1993
	
	
	
	Trefoil
	
	
	
	
	
	Lupinus angustifolius
	Lupins
	May_July
	o
	Williams et al., 1990; Fussell and Corbet, 1992; Williams, 1987
	
	
	L. albus
	Lucerneralfalfa
	
	
	
	
	
	Medicago sati¨a
	
	June_Oct
	f*
	Holm, 1966; Bohart, 1957; Free, 1993
	
	
	Melilotus albarofficinalis
	Sweet clover
	June_Oct
	q
	Fussell and Corbet, 1992
	
	
	Onobrychis ¨iciifolia
	Sainfoin
	June_Sept
	q*
	Richards and Edwards, 1988
	
	
	Phaseolus coccineus
	Runner bean
	
	f*
	Free, 1968a; Kendall and Smith, 1976
	
	
	Trifolium incarnatum
	Crimson clover
	June_July
	q
	Bohart, 1960
	
	
	Trifolium pratense
	Red clover
	May_Oct
	f*
	Teras, 1976; Macfarlane et al., 1983, 1991; Gurr, 1975; Free, 1993
	
	
	Trifolium repens
	White clover
	May_Oct
	f*
	Free, 1993; Michaelson-Yates et al., 1997; Fussell and Corbet, 1992
	
	
	Vicia faba
	Fieldrbroad
	June_July
	f*
	Free, 1993; Stoddard and Bond, 1987; Legeun et al., 1993;
	
	
	Vicia sati¨a
	bean
	
	q*
	Fussell et al., 1991
	
	
	
	Fodder vetch
	Apr_Sept
	
	Bohart, 1960; Free, 1993; Fussell and Corbet, 1992
	
	
	Vicia ¨illosa
	Hairy vetch
	June_Nov
	q*
	Wilson, 1929; Fussell and Corbet, 1992
	
	
	Pisum sati¨um
	Peas
	
	q
	Free, 1993
	
	Bumblebee Risk Assessment
	149
Table 1.	Crop species known to be used by Bombus spp.
	
	
	Common
	Flowering
	Visits by
	
	Family
	Species
	name
	season
	Bombusb
	References
	
	
	
	
	
	
	Linaceae
	Linum usitatissimum
	Linseedrflax
	June_Oct
	o
	Williams, 1988; Gubin, 1945
	Rosaceae
	Fragaria x ananassa
	Strawberries
	April_July
	o*
	Free, 1968b, 1993; Wilson, 1929; Nye and Andersen, 1974
	
	Malus sp.
	Apples
	March_May
	o*
	Free, 1993; Wilson, 1929; Kendall and Solomon, 1973; Mayer, 1983
	
	Prunus sp.
	Cherriesrplums
	April_May
	o*
	Wilson, 1929; Walton, 1927; Bhalla et al., 1983
	
	Pyrus sp.
	Pears
	April
	f
	Emmett, 1971
	
	Rubus fruticosus
	Blackberry
	May_Nov
	f*
	Fussell and Corbet, 1992; Wilson, 1929; Gyan and Woodell, 1987
	
	Rubus idaeus
	Raspberry
	May_Aug
	f*
	Willmer, 1994
	Solanaceae
	Capsicum annuum
	Sweet pepper
	
	f
	Free, 1993; Griffiths and Roberts, 1996; Shipp et al., 1994
	
	Lycopersicon esculentum
	Tomato
	May
	f*
	Free, 1993; Pinchinat et al., 1978; Kevan et al., 1993
	
	Solanum tuberosum
	Potato
	
	q
	Free, 1993; Sanford and Hannemann, 1981; Batra, 1993
a Denotes crops which do not produce flowers during commercial production. They will only be used by bumblebees when grown for seed.
b Frequency of visits:
f sfrequent: This indicates that bumblebees have been recorded, by at least one observer, to be frequent visitors. ‘Frequent’ is defined as )0.1 beesrm2. This threshold is arbitrary. It is often not comparable between different studies, due to variations in the densities of plants, flowers and other pollinators, and in the timing of fieldwork. However, the threshold, based on the findings of various studies, is felt to be useful in this context because it highlights those crops on which bumblebee populations might be seriously at risk from pesticides.
o soccasional: This indicates that no records have been found of bumblebees at )0.1 beesrm2 , under any circumstances.
qsunknown: This means that bumblebees have been recorded visiting the species, but no quantitative records of their density have been found.
* sknown to be visited by long-tongued bumblebees species	Bombus hortorum and B. pascuorum in the UK..
	150 Thompson and Hunt
Bumblebee Risk Assessment	151
Table 2.	The use of common arable weeds by bumblebees
	
	
	
	Number of
	
	
	
	
	
	studies
	
	
	
	Common
	
	recording
	
	
	
	name
	
	bumblebee
	
	
	
	frequency
	Flowering
	visits
	Do bumblebees show a preference for this
	
	Species
	rank .
	season
	total of 12.
	family or species?
	
	
	
	
	
	
	
	Stellaria mediaa
	Chickweed 1.
	All year
	1
	No
	
	Veronica sp. mostly
	Speedwells 2.
	All year
	2
	Family_Scrophulariaceae_exhibits bee pollination
	
	V. persica.a
	
	
	
	syndrome. These flowers are blue, but they are small
	
	
	Mayweeds 3.
	
	
	and almost actinomorphic, not typical of the family.
	
	Matricariar
	
	April_Oct
	3
	Family_Compositae_contains some bee pollinated
	
	Tripleurosperum
	
	
	
	flowers, but these species, with tiny yellow
	
	spp.
	
	
	
	disc florets, are not of that type.
	
	mostly T.
	
	
	
	
	
	maritima.a
	
	
	
	
	
	Capsella bursa-
	Shepherd’s
	All year
	1
	No
	
	pastorisa
	purse 4.
	
	
	
	
	Polygonum sp.
	Knotgrassrreds
	June_Nov
	2
	No
	
	mostly P.
	hank, etc. 4.
	
	
	
	
	a¨iculare.
	Charlock 4.
	
	
	
	
	Sinapis ar¨ensisa
	
	April_Oct
	1
	No
	
	Viola ar¨ensisa
	Field pansy 4.
	April_Nov
	3
	Yes. Family_Violaceae_typical of bee pollinated
	
	Galium aparinea
	Cleavers 4.
	
	
	flowers.
	
	
	
	May_Sept
	0
	No
	
	Con¨ol¨ulus
	Field bindweed
	June_Sept
	7*
	Yes. Large, nectar-rich trumpet shaped flowers. This
	
	arvensis
	5.
	
	
	species is perennial. Important to B. lucorum,
	
	
	
	
	
	B. terrestris and other
	
	
	
	
	
	short-tongued species.
	
	Lamium purpuremr
	Redrhenbit
	All year,
	9*
	Yes. Family_Labiatae _is typical of bee pollinated
	
	amplexicaulea
	deadnettle 5.
	March_Oct
	flowers. Nectar-rich,zygomorphic, purple.
	
	
	
	
	
	Especially important to B. Pascuorum.
	
	Papa¨er rhoeas
	Common
	June_Oct
	2
	Yes. Open flowered, pollen abundant species. Most
	
	
	poppy 5.
	
	
	likely to be used by B. lucorum.
	
a Weed species that was amongst the top 10 broadleaf weeds found in cereal crops Firbank, 1990..
This table lists the weed species that emerged as the most commonly found and gives information about the use of each of those species by bumblebees. References used in this review are given in the text. The plant species are ranked according to the number of papers reviewed in which they were listed as common weeds. Eleven studies were considered Teras, 1985; Fussell and Corbet, 1992; Patten et al., 1993; Fussell and Corbet, 1991; Walton, 1927; Prys-Jones, 1982; Saville, 1993; Barrow, 1983; Dramstad and Fry, 1995; Fussell and Corbet, 1993.. Any species that has been recorded by at least one author to be ‘important’ to bumblebees is indicated *. in the table.
Bumblebees have been recorded to forage on all but one of the ten species or groups of species identified as the most common arable weeds in the UK. Four of the weed species_field pansy Viola ar¨ensis., field bindweed Con¨ol¨ulus ar-
· ensis., red deadnettlerhenbit Lamium sp.., and common poppy Papa¨er rhoeas._exhibit fea-tures common to bumblebee pollinated flowers. Two of these _ red deadnettles and field bindweed_are considered to be important to
certain types of bumblebeein the seasonal suc-cession of forage plants.
Ellenberg 1988. lists red deadnettle and hen-bit Lamium purpureum and L. amplexicaule. as characteristic species of ‘summer crop weed com-munities,’ which indicates that they are more likely to be found growing under crops that are spring planted, as opposed to under winter planted cereals. Field bindweed Con¨ol¨ulus ar¨ensis., however, is not listed as characteristic of any
152 Thompson and Hunt
particular arable weed community, suggesting that it is ubiquitous.
2.2.	Field margin and hedgerow flowers
The flora of field margins and hedgerows is much more variable than the arable weed flora. It can depend on land use, both current and historical, soil type, boundary feature and climate and it is also drawn from a far greater pool of species which are able to thrive in the field margin habi-tat Barr et al., 1996; Mountford et al., 1994.. Therefore, it was not possible to treat the survey of bumblebee use in the same way as for arable weeds. A short list of ‘most common’ species, apart from being very difficult to compile, would be unlikely to represent any given field margin, verge or hedge. Therefore, those plant species found in field margins that are known to be frequently 1 visited by bumblebees have been listed. Farmers with these species in their margins or hedgerows risk causing damage to bumblebees if they allow spray drift, or spray pesticides in the field margins or hedgerows.
Fussell and Corbet’s 1992. national survey of flowers used by bumblebees was used as a basis for compiling a list of field margin and hedgerow flowers. Flowering plant species found in field margins and hedgerows were extracted from the lists for each colour group. These species are shown in Table 3 below, along with which bum-blebee colour group s . they are frequently used by. Woody species that make up hedges, such as hawthorn, have not been included in this list, because the flowers are less likely to be exposed to pesticides.
The studies of flowers visited by bumblebees used in the previous section Acker and Lutman, 1995; Barr et al., 1996; Clarke et al., 1993; Fir-bank, 1990; Glasgow et al., 1976; Gwynne and Murray, 1985; Knott et al., 1995; Lutman et al., 1995; Marshall, 1989; Mortimer, 1990; Olgivy et al., 1993; Pallutt, 1993; Watkinson and Bo, 1993. have also been reviewed with respect to field margin species common in the UK. Plant species which have been suggested to be important to bumblebees by other authors are indicated *.. If they are species not mentioned in the Fussell and Corbet survey, they are also listed in the table.
2.2.1. Results_field margin and hedgerow flowers of importance to bumblebees. Thirty-one plant species are listed Table 3., all of which are common in the field marginrhedgerow in at least some parts of the country Stace, 1997.. They are all perennial or biennial, with the exception of borage, and some species of vetch and geranium. All of them are considered important forage for bumblebees and are visited frequently by at least one of the colour groups. The only surprising inclusions in the table are angelica and hogweed, both of which are in family Umbelliferae and have relatively small, white flowers.
It must be pointed out that while these thirty-one species are used a lot by bumblebees, there are many other species that are also used by bumblebees.
One of the species in Table 3, Lamium album, white deadnettle, is of particular importance. It is an early flowering species and it is used exten-sively in the early Spring by foraging queens of long-tongued species Fussell and Corbet, 1992; Prys-Jones, 1982..
3.	Bumblebee ecology and activity patterns
There are 19 species of true bumblebee Bombus sp.. present in the UK for further information see Williams 1985., Prys-Jones and Corbet 1991... Bumblebee diversity in Britain has de-clined in recent decades and only six species remain widespread and common: Bombus lapidar-ius, B. lucorum, B. terrestris, B. pratorum, B. pascuorum and B. hortorum the first four are short-tongued species and the latter two long-tongued, see Table 4 for common names .. These six species will be considered in detail in this section, along with two other species which are known to visit crops in the UK_B. ruderarius and B. ruderatus. However, it must be considered that all the bumblebee species have similar habits. As members of the same genus, they have similar physiological, morphological and behavioural characteristics. Because the length of their season exceeds that of most flowering plant species, they are almost all generalists in terms of forage plants.
Most species are likely to be found close to arable farmland, which accounts for approxi-
Bumblebee Risk Assessment	153
Table 3.	Field marginrhedgerow flowers important to bumblebees
	
	
	
	Bumblebee colour groupsr
	
	
	
	species showing high
	
	
	Flowering
	preference for this
	Flower species
	Common name
	season
	speciesa
Species indicated by the national survey Fussell and Corbet, 1992. to be important:
	Borago officinalis
	Borage
	May_Sept
	2BWT
	Campanula spp.
	Bellflowers
	June_Sept
	BBRT
	Centaurea spp.*
	Knapweeds
	June_Sept
	2BWT; BBRT; BRT
	CirsiumrCarduus spp.*
	Thistles
	June_Sept
	BBRT; BRT
	Digitalis purpurea*
	Foxglove
	June_Sept
	3BWT
	EpilobiumrChamerion spp.*
	Willowherbs
	June_Aug
	2BWT
	Geranium spp.
	Cranesbills
	June_Sept
	BRT
	Lamium album*
	White deadnettle
	March_Nov
	Br; 3BWT
	Lonicera spp.
	Honeysuckle
	June_Oct
	3BWT
	Lotus corniculatus*
	Birdsfoot trefoil
	May_Sept
	BBRT
	Ranunculus spp.
	Buttercups
	April_Oct
	BBRT
	Rubus fruticosus*
	Bramble
	May_Nov
	2BWT
	Senecio jacobaea
	Ragwort
	June_Nov
	BRT
	Stachys spp.*
	Woundworts
	April_Oct
	Br; 3BWT
	Symphytum spp.*
	Comfrey
	May_June
	2BWT; BRT; Br
	Trifolium pratense*
	Red clover
	May_Oct
	Br; 3BWT
	Trifolium repens*
	White clover
	May_Oct
	BBRT
	Vicia spp.*
	Vetches
	April_Sept
	Br
	Species considered important by other authors:
	
	
	Angelica syl¨estris Saville, 1993;
	Angelica
	July_Sept
	
	Dramstad and Fry, 1995.
	
	
	
	Ballota nigra Fussell and Corbet,
	Black horehound
	June_Sept
	
	1991; Saville, 1993.
	
	
	
	Clinopodium ¨ulgare Saville, 1993.
	Wild basil
	July_Sept
	
	Dipsacus fullonum Saville, 1993.
	Teasel
	July_Sept
	
	Galeopsis tetrahit Saville, 1993.
	Hemp nettle
	July_Sept
	
	Heracleum sphondylium Fussell and
	Hogweed
	April_Nov
	B. terrestris rlucorum Fussell
	Corbet, 1991.
	
	
	and Corbet, 1993.
	Lamiastrum galeobdolon Barrow,
	Yellow archangel
	May_June
	
	1983.
	
	
	
	Linaria ¨ulgaris Saville, 1993.
	Toadflax
	June_Oct
	
	Knautia ar¨ensis Saville, 1993.
	Field scabious
	June_Oct
	B. pascuorum Brian, 1957.
	Solanum dulcamara Brian, 1957;
	Bittersweet
	May_Sept
	
	Saville, 1993.
	
	
	
	Solidago ¨irgaurea Saville, 1993.
	Golden-rod
	June_Sept
	
	Taraxacum officinale Barrow, 1983.
	Dandelion
	April_June
	
	Teucrium scorodonia Williams,
	Wood sage
	July_Sept
	B. terrestris, B. lapidarius
	1985.
	
	
	Williams, 1985.
a Colour groups: 2BWTs2-banded white tails; 3BWTs3-banded white tails; BBRTsblack-bodied red tails; BRTsbanded red tails; Br sbrowns.
Species in the table are presented in alphabetical, rather than rank order, because the latter is different for each colour group. Species in the first half of the table are all from the ‘top ten’ for the colour groups mentioned in Fussell and Corbet’s list 1992.. Species only mentioned in other surveys are shown in the second half of the table. Species considered important by at least one other author are indicated *. see text for references ..
Table 4.	Phenologies of eight bumblebee speciesa
	
	
	
	
	
	Peak worker
	
	
	
	
	Emergence of
	Emergence of
	
	numbers and
	
	
	Species
	Queen emergence
	first workers
	first males
	Total season
	size a
	Regional differences
	
	
	
	
	
	
	
	
	
	B. terrestris:
	Late Feb_early Apr
	Late Apr_
	Mid-June_
	End Feb_end Oct
	End July
	Whole cycle earlier in WL; much later
	
	buff-tailed
	
	end Junemid-July
	long cycle .
	200_300
	in Kent colony over early in SW
	
	bumblebee
	Early Mar_mid-Apr
	Early May_
	Mid-June_
	Early Mar_end
	Mid-July_Aug
	by mid-Aug in 1980.
	
	B. lucorum:
	
	
	
	
	
	First queens to emerge in SW; whole cycle
	
	common
	
	early July
	start July
	Sept
	200_300
	later in Kent
	
	white-tailed
	
	
	
	long cycle.
	
	
	
	bumblebee
	
	
	
	
	
	
	
	B. pratorum:
	Mid-Mar_mid-Apr
	Mid_late
	Early May_
	Mar_mid-Sept
	Early_mid June
	First queens to emerge in C1
	
	meadow
	
	April
	end July
	short cycle .b
	usually approx 30,
	Colony over in July in WL
	
	bumblebee
	
	
	
	
	sometimes up to
	2nd worker peak in Aug, C2b
	
	
	
	
	
	
	200
	
	
	B. ruderarius:
	Mid-Mar_late May
	Late May_
	Mid_late July
	Mar_early Sept
	Early July_late
	Last species to emerge in SW, but colony
	
	Small red-
	
	end June
	
	short cycle .
	Aug
	over in Aug
	
	tailed bumblebee
	
	
	
	?
	
	
	B. lapidarius:
	Mid-Mar_start May
	Mid-May_
	Early June_
	Mar_mid-Sept
	Mid June_Aug
	Whole cycle later in Kent, worker peak
	
	Large red-
	
	mid-June
	early Aug
	medium cycle .
	200_300
	Aug in C1, June in C2
	
	tailed bumblebee
	
	
	
	
	
	
	B. ruderatus:
	Early May
	Late June
	Early July
	?
	?
	Only recorded in Kent
	
	Ruderal
	
	
	
	
	
	
	
	bumblebee
	
	
	
	
	
	
	
	B. pascuorum:
	Late Mar_mid-May
	Early May_
	Late June_
	Late-Mar_mid-Oct
	Aug
	Whole cycle early in WL, late in Kent,
	
	Common
	
	mid-June
	mid July
	long cycle .
	Less than 200
	length of time queens_first
	
	carder bee
	
	
	
	
	
	workers variable, longest in C1, but
	
	
	
	
	
	
	
	males emerging similar time at all sites
	
	B. hortorum:
	End Mar_mid June
	Early May_
	End May_
	End Mar_mid-Oct
	Late May_late
	Whole cycle much later in Kent,
	
	garden
	
	mid-July
	mid July
	short cycle .b
	July
	early queen emergence in WL and S,
	
	bumblebee
	
	
	
	
	30_80
	season ends late July_early
	
	
	
	
	
	
	
	Sept in K, S and WL
	
The information in this table has been collected from several detailed studies Prys-Jones, 1982; Saville, 1993; Williams, 1985; Goodwin, 1995; Barrow, 1983._these were carried out in Cambridgeshire C1 and C2., Kent K., West London WL., south Wales SW. and north-west Scotland S. respectively. Species are presented in approximate order of emergence in Spring, although this does not entirely agree between different studies. Dates given are the range across all of the studies for each species. Any notable discrepancies between the different studies are described in the final column, Regional Differences.
a Data from Prys-Jones and Corbet, 1991; Alford, 1975.
b Signifies evidence for bivoltinism_for example, where a short cycle species has been recorded late in the year, but not at all sites, or where a second peak in worker numbers has been recorded late in the season.
	154 Thompson and Hunt
mately 20% of land cover across the UK MAFF, 1997.. Even Bombus monticola, which is consid-ered a moorland species Alford, 1975. has been shown to rely on arable field margin and verge flowers at certain times of year Yalden, 1982..
3.1.	Natural history
All bumblebee species form colonies which are small in comparison to honeybee colonies. A colony of several hundred workers is considered large in bumblebees Prys-Jones and Corbet, 1991. compared to a full colony size of around 30,000 individuals, for honeybees Seeley, 1985..
The life cycle of bumblebees in temperate re-gions differs from that of honeybees in that only the queens over-winter. The rest of the colony_ workers and males_only survive for a single sea-son. In Spring, mated queens emerge, feed and establish a new colony alone. They must collect pollen and nectar to feed their first batch of worker larvae, as well as feeding themselves. Of-ten the queen works alone for more than a month, before her first workers appear.
New gynes reproductive females . and males are produced in late summer to early autumn. Once emerged, males do not return to the nest. They forage for themselves, ‘‘patrol’’ for mates and usually spend the night on flower heads. At the end of the season, mated queens search for sites to hibernate underground. The rest of the colony dies.
Bumblebees, then, are much more vulnerable than honeybees, in that there is a month every year when the entire population depends upon the success of the queens in establishing colonies. This makes the colony particularly susceptible to pesticides applied early in the year. In addition, as they have smaller colonies, a single bumblebee worker is more important to the survival of the colony than a single honeybee worker.
3.2.	Phenology
Phenology is the annual, or seasonal, pattern of activity, which is specific to each bumblebee species. It varies according to parameters such as colony cycle length and forage requirements, as well as external factors like climate. Prys-Jones 1982. and Goodwin 1995. both showed that the
Bumblebee Risk Assessment	155
timing of queens emerging from hibernation cor-relates with temperature_a maximum soil tem-perature of 6_9_C in the former study and air temperature of 11_C in the latter. In general, the bumblebee season runs from mid-March to mid-October, with a peak in numbers during the sum-mer. Table 4 presents available information on the timing of the annual cycle, for the eight species of bumblebee that have been recorded to visit crops in the UK.
The order in which the species emerge from hibernation is relatively predictable, although there was disagreement between the studies over the order of the first three species to emerge. It is worth noting that the two long-tongued species in the list B. pascuorum and B. hortorum. are the last species to emerge. This means that foraging queens of these species in the process of founding colonies can be vulnerable to the effects of pesti-cides later in the year than those of other species.
3.3.	Diel acti¨ity patterns
The foraging and flying activity of bumblebees during the day has been recorded by many au-thors. Normally, the pattern observed is that the number of foragers peaks in the early morning and evening, with a drop in numbers in the mid-dle of the day Plowright and Laverty, 1984; Al-ford, 1975.. Peaks are usually recorded before 10:00h and after 16:00h. Bumblebees also tend to start foraging earlier in the day than honeybees, and finish later in the evening Fussell and Cor-bet, 1991; Corbet et al., 1993.. This is very differ-ent from the activity pattern for honeybees, in which the number of foragers peaks in the middle of the day.
Two reasons are thought to combine to ac-count for this difference in foraging activity be-tween bumblebees and honeybees_the effects of ambient temperature and the effects of exploita-tive competition with honeybees.
3.3.1. Temperature. Most bumblebees are able to fly at lower air temperatures than honeybees Corbet et al., 1993; Lundberg, 1980; Stone and Willmer, 1989.. This is because bumblebees have better thermoregulatory ability and they are able to warm up their wing muscles to a temperature considerably higher than ambient Heinrich,
156 Thompson and Hunt
1993.. These data confirm that generally, bumble-bees can be active at lower temperatures than honeybees, with the possible exception of B. lapi-darius. Bumblebees may be limited by higher tem-peratures, because they are large and black and at risk of overheating Willmer, 1983.. This may partly explain the observed drop in foraging activ-ity during the middle of the day, in the summer months.
3.3.2. Competition from honeybees. A second, and complementary, explanation for the diel ac-tivity pattern of bumblebees is that they confine their foraging to times when honeybees are less numerous. Honeybees can be present in very large numbers, especially since they are known to com-municate the whereabouts of resources to one another Seeley, 1985. and they often congregate at a good foraging site. They can deplete the quantity of nectar in each flower to a level which may be below that required for bumblebees to forage profitably.3.4. Bumblebee foraging beha¨iour in relation to that of honeybees
Bumblebees almost always forage faster than honeybees. They generally spend a shorter time per flower and visit more flowers per minute than honeybees Williams, 1997.. For both types of bee, flower handling time can be expected to increase as the quantity of available nectar in each flower increases Williams, 1997.. Long-tongued bumblebees tend to forage even faster than short-tongued bumblebees Alford, 1975..
Unlike honeybees, which use information from other foragers, bumblebees learn where to forage by their own initiative Prys-Jones and Corbet, 1991.. They are not known to communicate infor-mation about food sources and each individual bases foraging decisions on its own experience Plowright and Laverty, 1984. . Bumblebees ex-hibit a behaviour pattern called ‘trap-lining,’ in which an individual worker visits the same se-quence of flowers on each foraging trip Saville, 1993.. They are less ‘flower constant’ faithful to one plant species . than honeybees Plowright and Laverty, 1984. and they regularly sample flowers of minority plant species, or newly opened flowers Heinrich, 1979..
Bumblebees are thought not to forage very far away from their nests. Many authors have sug-gested that they will not exceed 250 m from home Saville, 1993.. However, in a study of bumblebee dispersal in an arable habitat, Saville 1993. dis-covered at least one individual more than 300 m from the nest and found many marked bees could not be located in the study area once they had left the nest to forage. The exact range of forag-ing bumblebees, then, remains uncertain.
The quantity of nectar which a bumblebee drinks from a flower has been investigated by Prys-Jones 1982.. He tempted bees into a glass tube, in the field and fed them sugar solution from a syringe. Only B. hortorum could not be persuaded to participate, and so for that species, measurements were made by feeding the colony in a nest box. Uptake rate and total volume of sugar solution imbibed were found to vary accord-ing to body weight and concentration of the sugar solution.
The ranges of nectar uptake rate and total quantity taken are given in Table 5 below for the four species studied. The uptake rate is positively correlated to body size, such that larger bees can drink faster _in general, a doubling of body weight led to a 30_40% increase in uptake rate. In real flowers long-tongued bees have been recorded to be faster drinkers than short-tongued bees of similar body size Harder, 1983..
The way in which these observations affect the potential exposure of bumblebees to pesticides depends on whether pesticides become dissolved in the nectar of flowers, and if so, exactly how that happens. If pesticides enter nectar immedi-ately by direct contact with the flowers, then to spray in humid conditions may be a disadvantage to bumblebees because they may be drinking faster at the time. On the other hand, if pesticides are absorbed by the plant and secreted in the nectar, then bumblebees will receive a higher dose if the nectar is more concentrated, in dry conditions.
3.5.	Preferences in flower use
3.5.1. General bumblebee preferences. There are several sets of flower characteristics particularly associated with bee pollination bee pollination syndromes. Proctor et al., 1996. including one
Bumblebee Risk Assessment	157
Table 5.	Nectar uptake in different bumblebee species
	Bumblebee
	Nectar uptake
	Total volume
	Range of unfed
	speciesrcaste
	rate _lrs .
	nectar taken _l .
	mass mg.
	
	
	
	
	B. hortorum:
	
	
	
	workersrgynes
	0.9_3.0
	
	
	B. pascuorum: workers
	
	36.0_65.1
	74_165
	B. pratorum
	
	
	
	Workers
	0.3_1.8
	
	83_160
	Gynes
	0.5_3.1
	
	325_425
	B. terrestris
	
	
	
	Workers
	0.5_2.0
	41.1_111.9
	109_300
	Gynes
	
	104.0
	
	A. mellifera
	
	50 a
	100b
This table shows the uptake rates and total volumes of nectar of varying concentrations, taken by different bumblebee species. It also gives the range of mass measured, to give an indication of sizes modified from Prys-Jones 1991...
a Crane 1990..
b NBU data.
peculiar to bumblebees. However, bumblebees are opportunists and will visit a wide variety of species including some families that characterise differ-ent pollination syndromes such as Umbelliferae largely fly-pollinated . Corbet et al., 1991.. Bum-blebees tend to pollinate the larger flowers Proc-tor et al., 1996.. In addition, many authors have shown that bumblebees prefer to forage on perennial flowers, as opposed to annual species Fussell and Corbet, 1992; Prys-Jones, 1982; Sav-ille, 1993; Dramstad and Fry, 1995.. This prefer-ence can be largely explained by the fact that perennial species often produce more nectar per flower or per plant than annual species Fussell and Corbet, 1992..
Some bumblebees have been observed to col-lect pollen from non-insect pollinated plant species, such as grasses or salt-marsh plants Pojar, 1973.. At certain times of year, when reproductives are being reared, for example mid-summer., it is possible that pollen is a limit-ing resource for bumblebees, forcing them to move to plants that produce copious pollen Plowright and Laverty, 1984..
3.5.2. Interspecific differences in flower preference. It has often been said that bumblebee species are able to coexist because they have different flower preferences and different species of bee forage on different species of plant. This may be at least partly attributable to differences in tongue length between the species.
Preferences are shown in Table 6 below for each of the six common species only these species have been studied in any detail .. The table also gives estimates for the average tongue length of workers of each species, as calculated by Saville, from a variety of sources Saville, 1993.. This serves to demonstrate the morphological differ-ence which is partly responsible for species spe-cific flower preferences.
4. Toxicity and repellency of pesticides to bumblebees
4.1. Toxicity, repellency and sub-lethal effects data
Van der Steen 1994. showed that the acute con-tact and oral toxicity of dimethoate is correlated with the size of the bumblebee B. terrestris.. This reported the toxicity of dimethoate to five size classes of bumblebees ranging from 0.162 g to 0.297 g. When corrected for weight bees in the mid-range 0.168_0.285 g. have similar LD50 val-ues, 33_37 _grg bee, but small bees 0.162 g. have a lower LD50 of 25 _grg bee and large bees
0.297 g. have a higher LD50 of 44 _grg bee. Therefore although correcting for size can reduce
the variability in the mid-range of size signifi-cantly smaller, presumably younger, and signifi-cantly larger, probably older, bees have differing LD50 s in terms of weight. It is therefore impor-tant that toxicity data for bumblebees are quoted
158 Thompson and Hunt
Table 6.	Suggested flower preferences and tongue lengths for six common bumblebee species
	
	
	
	
	Average tongue
	
	Bumblebee species
	Flower preferences
	length mm.a
	
	
	
	
	
	
	
	B. lucorum
	v
	Wide range of short tubed and open flowers
	7.06
	
	
	
	
	
	
	
	
	Brian, 1957.
	
	
	
	
	v
	Often collecting only pollen Brian, 1957;
	
	
	
	
	
	
	
	
	
	Fussell and Corbet, 1992; Macdonald, 1998.
	
	
	
	v
	Clustered flowers for nectar only Prys-Jones,
	
	
	
	
	
	
	
	B. pratorum
	
	1982.
	
	7.60
	
	
	v
	Open flowers, prefers to forage in shady
	
	
	
	
	habitats Brian, 1957.
	
	
	
	v
	Makes good use of pendulous flowers_due to
	
	
	
	
	
	
	
	
	
	small size and agility? Prys-Jones, 1982;
	
	
	
	
	Macdonald, 1998.
	
	
	
	v
	
	b
	
	
	
	
	Actinomorphic flowers Fussell and Corbet,
	
	
	
	
	
	
	
	
	
	1992.
	
	
	
	B. lpidarius
	v
	Composites or clustered flowers Fussell and
	7.77
	
	
	
	
	
	
	
	
	Corbet, 1992; Prys-Jones and Corbet, 1991;
	
	
	
	
	Prys-Jones, 1982.
	
	
	
	v
	Preference for yellow flowers Fussell and
	
	
	
	
	
	
	
	
	
	Corbet, 1992.
	
	
	
	
	v
	Flowers facing upwards Fussell and Corbet,
	
	
	
	
	
	
	
	B. terrestris1992.
	
	8.29
	
	
	v
	Clustered flowers for nectar only Prys-Jones,
	
	
	
	
	1982.
	
	
	
	
	v
	Upwards facing flowers Prys-Jones, 1982.
	
	
	
	
	
	
	
	B. pascuorum
	v
	Flowers of intermediate corolla length; prefers
	8.94
	
	
	
	
	
	
	
	
	sheltered habitats Brian, 1957.
	
	
	
	v
	Horizontal facing flowers, like Labiates Prys-
	
	
	
	
	
	
	
	
	
	Jones and Corbet, 1991.
	
	
	
	v
	Prefers zygomorphic flowers Fussell and
	
	
	
	
	
	
	
	
	
	Corbet, 1992; Macdonald, 1998.
	
	
	B. hortorum
	v
	Flowers with long corolla tubes Brian, 1957;
	12.87
	
	
	
	
	
	
Fussell and Corbet, 1992.
	v
	Deep, nectar rich flowers Prys-Jones and
	
	
	
	
	
	Corbet, 1991.
	
	v
	Zygomorphic flowers, especially with spurs
	
	
	
	
	
	Fussell and Corbet, 1992.
	
a Prementum and glossa together.
b Radially symmetrical.
in terms of weight of the bees tested as, unlike honeybees, their weight can vary significantly be-tween individuals.
Table 7 shows the contact and oral toxicity of a range of pesticides to honeybee and bumblebee species both in terms of per bee and per g bee. There are very limited data for bumblebee species other than B. terrestris although the data avail-able shows the toxicity on a weight basis to be similar. It can be seen that the toxicity of the pesticides, for which data are available, are gener-ally lower to bumblebees than honeybees when expressed on a weight basis. However, it should also be remembered that the data are limited in
terms of number and type of insecticide. There is a need to increase the amount of toxicity data available for bumblebees in order to support the assumption that they are less susceptible than honeybees.
5. Likelihood of exposure of queens and workers to pesticides
Overlaying the annual patterns of emergence of queens and workers for the most common bum-blebee species with seasonal uses of insecticides on crops shows that there are a number of classes
Table 7.	Contact and oral toxicity of a range of pesticides to honeybee and bumblebee species 24 hr unless otherwise stated.
	
	Contact LD50
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	Contact LD50
	
	
	
	A. mellifera
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	_g airbee
	
	Oral LD50
	Contact LD50
	
	Oral LD50
	
	Contact LD50
	B. agrorum
	Oral LD50
	
	Pesticide
	_ g rg bee .a
	A. mellifera
	B. terrestris
	
	
	
	B. terrestris
	B. lucorum
	Pascuorum.
	B. lapidarius
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	Acephate
	
	
	
	
	
	0.2
	c
	( )
	
	
	
	
	
	
	
	
	135.5
	c (
	
	)
	
	
	
	
	
	
	
	
	
	
	
	
	2.0
	
	
	
	
	
	
	
	
	
	645
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	3.93, 72 hr
	c
	( )
	
	
	
	Alpha q
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	19
	
	
	
	
	0.03e
	(0.3)
	
	0.06e (0.6)
	0.17i (0.81)
	
	
	
	0.52 g
	2.5.
	
	
	
	
	
	cypermethrin
	0.05g (0.5)
	
	
	
	
	0.15, 72 hr i
	(0.71)
	0.36, 72 hr g
	1.7.
	
	
	
	Chlorpyrifos
	0.059
	g
	(
	)
	
	
	
	2.39
	i
	(
	)
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	0.59
	
	
	
	
	
	11.4
	i
	(
	
	)
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	1.58, 72 hr
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	e
	(
	)
	
	
	
	
	
	
	7.5
	
	
	b
	(
	)
	
	
	
	Deltamethrin
	0.05
	
	
	
	
	
	
	
	0.9, 48 hr Decis CE
	0.6 Decis
	
	
	
	
	
	
	
	
	
	0.5
	
	
	
	
	
	
	
	2.7
	
	
	
	
	
	
	
	
	
	
	
	
	in acetone
	f
	( )
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	4.3
	
	
	
	
	
	
	
	
	Demeton methyl
	0.5
	h
	(
	)
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	Workers 1-2
	h
	Workers 1-3
	h
	(
	)
	
	
	
	
	
	5.0
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	8-25
	
	
	
	
	
	g (
	)
	
	
	
	
	i (
	)
	
	
	
	
	
	
	
	
	Queens 6-24 h
	Queens 10-24h
	
	
	Demeton S
	0.26
	
	
	
	
	
	3.27
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	2.6
	
	
	
	
	15.6
	i
	(
	)
	
	
	
	
	
	
	
	
	
	
	
	
	
	methyl
	
	
	
	
	
	
	
	
	2.68, 72 hr
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	13
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	Dimethoate
	0.4d (4.0)
	0.12, 24
	i ( )
	4.1-13c (19-62)
	4.7, 24-72 hr i
	Workers 2-5 h
	Workers 0.5-2h (4-17)
	
	
	
	
	
	
	
	& 48 hr
	
	
	
	
	
	
	(
	
	
	)
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	1.2
	
	
	
	
	
	22
	
	
	
	
	
	
	
	
	
	
	
	
	0.12, 24
	
	
	
	
	4.8, 24-72 hr i (23)
	
	
	
	
	
	
	Queens 5-20 h
	Queens 1-5h
	
	
	
	
	
	& 48 hr g (1.2)
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	h
	(
	)
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	0.1
	
	
	1.0
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	Disulfoton
	5.0h (50)
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	Workers 2-10 h
	Workers 1-4h (8-33)
	
	
	
	
	
	
	
	
	c ( )
	
	
	
	
	
	
	c
	
	( )
	
	
	Queens )40 h
	Queens 5-10h
	
	
	
	Methomyl
	
	
	
	
	
	0.08
	
	
	
	
	
	
	3.2
	
	,
	
	2.6,
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	0.8
	
	
	
	
	
	
	
	
	
	15
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	72 hr
	c
	( )
	
	
	
	
	
	
	
	
	0.54g (5.4)
	
	
	
	
	
	
	
	
	
	
	12
	
	
	
	
	
	
	
	Oxydemeton
	
	
	
	
	
	
	
	
	
	0.75 Metasystox
	
	
	
	
	
	
	methyl
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	b
	(
	
	)
	
	
	
	
	
	
	
	
	
	0.1e
	(1.0)
	0.03e (0.3)
	
	0.81i (3.9)
	
	
	
	R
	3.6
	
	
	
	
	
	
	
	
	Permethrin
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	0.3h (3.0)
	
	
	
	0.82, 72 hr (3.9)
	
	
	
	
	
	
	Workers 1-2 h
	Workers 1-2h (8-17)
	
	Phorate
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	Queens 6-23 h
	Queens 1-5h
	
	
	
	
	Phosalone
	8.9g (89)
	
	
	
	
	
	
	
	
	60 b Rubitox (286)
	
	
	
	
	
	
	Pirimicarb
	)54 g
	
	
	
	
	
	
	
	
	
	8.5 Pirimor
	b
	
	
	
	
	
	
	
	
	()
	
	)
	
	
	
	
	
	
	
	
	Granulat
	
	( )
	
	
	
	
	
	
	
	
	
	
	540
	
	
	
	
	
	
	
	
	
	
	40
	
	
	
	
	
	
2.78 c , 2.4, 48 hr c , 2.18,
72 hr c
	Bumblebee Risk
a Calculated av A. melliferas0.10 g, B. terrestris s0.21 g, B. pascuorums0.12 g; b Gretenkord and Drescher, 1993.; c Drescher and Geusen-Pfister, 1991.; d van der Steen, 1994.; e Inglesfield, 1989.; f Tasei et al., 1994.; g Greig-Smith et al., 1994.; h Stevenson and Racey, 1996.; i unpublished NBU data.
	Assessment
	159
160 Thompson and Hunt
of insecticide applied to crops during the periods when queens are emerging and establishing colonies.
Of particular concern is that pyrethroids can be applied to oilseed rape crops in full flower at times of day when, although honeybees are less active, bumblebees are likely to be foraging on the crop. There are recommendations that pyrethroid sprays are applied early morning or late evening to avoid honeybee exposure. How-ever, as shown above, these are the peak activity times for bumblebee species. Therefore direct overspray of foraging bumblebees by pyrethroids is likely to occur.
Although the application of these pesticides to the remainder of these crops is not of immediate concern, as there are label restrictions to prevent applications to flowering crops, there are likely to be flowering weeds around many of these crops which are attractive to bumblebees. Generally queens can be seen to emerge between February and mid-June depending on species and during this period many of the common weed species are in flower. Flowering weeds are unlikely to be visited by significant numbers of honeybees and therefore are likely to be overlooked by spray operators. Therefore the presence of flowering weeds in and around agricultural crops on which bumblebee species may forage is important in determining the exposure of bumblebees.
6. Quantitation of exposure of bumblebees relative to honeybees
6.1.	Oral exposure
Oral exposure by uptake of nectar from treated crops and oversprayed weeds depends on the number of flowers visited, the level of transfer of pesticide to nectar and the capacity of the honey stomach of the bee. In addition, oral exposure may occur during the collection of contaminated pollen and is related to the number of flowers visited. On average honeybees undertake 7_13 foraging trips per day Crane, 1990. whereas bumblebees undertake 17_27 foraging trips per day Alford, 1975.. Bumblebees visit an average of 2.5 more flowers per minute than honeybees and have a nectar carrying capacity per trip of up
to 112 _l compared to 50 _l in a honeybee. Therefore bumblebees have the potentialto take up to 5 times the level of contaminated nectar in a day as honeybees. There are no data available on the levels of pesticides in pollen.
6.1.1. Secretion of pesticides into nectar. There are two possible routes of exposure of bumble-bees to pesticides, through uptake of nectar into which the pesticide has been secreted or through contact with treated foliage or flowers. Systemic pesticides are most likely to occur in nectar and their concentrations depends on both the amount and method of nectar secretion. Investigations with dimethoate and carbofuran in Ajuga reprans, Brassica napus and Vicia faba showed an apparent selective transport of the insecticides into the nectar as the concentration in nectar often ex-ceeded that in the solution in which the excised flowers were exposed, i.e. it is more than passive movement with water Davis and Shuel, 1988..
Barker et al. 1980. showed that one day after application of dimethoate to an alfalfa crop at 340 ppm active ingredient residues of dimethoate in pollen were 0.5 ppm but were 16 ppm in nectar of uncovered florets and 5 ppm in the nectar from covered florets. After one week the residues in nectar had declined to 3 ppm and after 2 weeks 1 ppm. These values for dimethoate in the nectar of alfalfa are similar to the levels reported for fuschia, nasturtium and field beans by Lord et al. 1968.. Pesticides in nectar are not thought to harm honeybees as they are diluted with stores within the colony. Bumblebees do not build up such high levels of stores and therefore are likely to consume higher levels.
There are several reports that it is not only the truely systemic pesticides which can be detected in nectar but chemicals such as parathion can result in toxic nectar for up to 24 hours Jaycox, 1964.. Even systemic granular insecticides can penetrate sufficiently into nectar to kill bees Jaycox, 1964..
Lord et al. 1968. showed that 6 days after nasturtium and fuschia plants watered with 25 mg dimethoate resulted in 741 "259 ngrml dime-thoate in nasturtium nectar and 2890 "550 ngrml in fuschia nectar 0.05 ml nectar .. 50 mg dimethoate applied to compost in which field bean plants were growing resulted in 130_136
ngr6 mg in nectaries whereas 100 mg of phorate resulted in only 0.042_0.052 ngr6 mg in nec-taries.
6.1.2. Assessment of exposure. The transfer of pesticide to nectar may be by direct overspray or by translocation within the plant. Using the data of Barker et al. 1980. a 340 ppm ai dimethoate spray resulted in 16 ppm _grml. in nectar from uncover alfalfa florets, i.e. a 5% recovery. There-
fore with an oral LD50 of 4.7 _g airbee a bum-blebee must consume 293 _l nectar to obtain a
lethal dose and a honeybee 7.5 _l	LD50 0.12
· grbee .. If a flower contains 50 _l nectar this equates to 6 flowers visited for bumblebees and only 1 flower for a honeybee. Given that bumble-bees undertake twice as many foraging trips dur-ing a day and 2.5 times as many flower visits then the level of risk is similar for the two species when based on their behaviour and the toxicity of the pesticides.
6.2.	Contact exposure
Obviously, when it comes into direct contact with a pesticide, the residue on a bumblebee is greater than that on a honeybee due to its larger surface area. Assuming that there is not a significant difference in density of bee species then the weight of the species is related to surface area, i.e. bumblebees are approx twice the weight and twice the surface area of honeybees. Therefore differences due to size are taken into account in data relating toxicity as _grg bee and residues as
· grg bee. However, Bombus species tend to be more active foragers visiting more flowers per min on average 2.5 times . and undertaking ap-prox twice as many foraging trips in a day than Apis mellifera which will increase their exposure level, i.e. the exposure of bumblebees to residues on flowers may be up to 5 times that of honey-bees based on their behaviour.
6.2.1. Assessment of exposure. Koch and Weiber 1997. reported the results of a field study using a fluorescent tracer to assess the exposure of hon-eybees to pesticides applied to crops. The only route which could be assessed in this manner is contact exposure but it provides information
Bumblebee Risk Assessment	161
which may be extrapolated to other species with similar behaviour patterns. Assessments of the levels of tracer on bees returning to the hive were undertaken for 20_30 minutes after the spray application. The mean deposit per bee for approx. 100 bees sampled at 5 minute intervals over the 20_30 minutes was 1.62_20.84 ng in apple or-chards and 6.34_35.77 ng in Phacelia following applications at 20 grha. The maximum residue detected following the orchard application was 35 ng and following the Phacelia application 48 ng. This was despite the potential for prolonged ex-posure of bees in orchards to be greater than in arable crops due to the time required for spray application 20 minrha in fields, 75 minrha in orchards .. The residues detected declined to low levels less than 10 ng. within 30 minutes of finishing the application. Therefore, an applica-tion rate of 20 grha results in a mean residue of 18 ng per bee in Phacelia crops. Given the sur-face area of a bumblebee is approx. 2.5 times that on a honeybee, under similar circumstances the residue of tracer on the surface of a bumblebee would be in the order of 46 ng.
6.3.	Risk assessment
Risk assessments are routinely based on the dose available mgrkg. and the toxicity of the com-pound mgrkg.. However, for honeybees an em-pirical approach hazard ratio . has been devel-oped EPPO, 1993. based on the application rate of the pesticide g airha . and the toxicity to the bee _grbee .. These approaches are reviewed below together with other methods of assessing exposure which may be more readily adapted for use with bumblebees.
6.3.1.	Hazard ratio.	A hazard ratio	application
rate g airha.rLD50 _grbee . of -50 is used to define a pesticide as harmless to honeybees,
50_2500 as slight to moderately toxic and )2500 as dangerous to bees. An application rate of 15 g airha for alphacypermethrin gives a hazard ratio of 88 for bumblebees and 500 for honeybees. However, this use of a hazard ratio does not take into account the differences in the foraging be-haviour and thus the exposure of bumblebees.
162 Thompson and Hunt
6.3.2. Insect residue based data. Pesticide residues on large insects mgrkg. are calculated by Kenega as 2.7 times the application rate in kgrha. Therefore, at an application rate for al-phacypermethrin of 0.015 kgrha the residue on a large insect would be 0.041 _grg. Taking the toxicity of alphacypermethrin to bumblebees as 0.81 _grg bee this gives a dosertoxicity DrT. of
0.05 and a medium risk classification. The classifi-cation for honeybees would be DrT 0.13, a high risk classification.
6.3.3. Bee residue data. Using the data pro-duced by Koch and Weiber 1997. based on ac-tual residues following a field application of a marker to produce a DrT is another method for assessing the risk DrT.. At 15 g airha the Koch and Weiber 1997. data gives a residue of 35 ngrbee 0.175 _grg bee . for bumblebees and for alphacypermethrin a DrT of 0.22. For honeybees the same data gives a residue of 13.5 ngrbee 0.135 _grg bee . and DrT of 0.45 for alpha-cypermethrin. These are both high risk classifica-tions. This method allows the larger surface area of the bumblebee to be taken into account in the risk assessment. However, further data are re-quired to determine the scale of difference in contact exposure between bumblebees and hon-eybees based on their behaviour, e.g. number of foraging trips and number of flowers visited which would allow extrapolation of the risk assessment.
7. Possible role of pesticides in decline of bumblebees
7.1.	Non-Pesticide causes of bumblebee decline
As is the case with butterflies Longley and Sotherton, 1997. there are two main potential causes of bumblebee decline on farmland, a de-crease in suitablehabitats andr or the toxicologi-cal effects of insecticides.
The potential for indirect effects on bumble-bees through the loss of nectar producing plants, e.g. perennial weeds, as a consequence of herbi-cide use is significant. In addition the use of
nitrogen fertilisers can reduce floral diversity in field boundaries and thus essential nectar bearing plants. The removal of hedgerows and ploughing of headlands will also result in the loss of nest sites, hibernation sites, male patrolling and mat-ing sites.
7.2.	Pesticide effects
Bumblebees may be exposed to pesticides by di-rect overspray, spray drift or vapour drift through contact with treated flowering crops or small patches of flowering weeds. In addition there are a number of sublethal effects of pesticides which may impact on colonies from the effects of insect growth regulators on brood development to ef-fects of pyrethroids on memory and ability to return to the colony.
Bumblebees are likely to be more attracted than honeybees to small, isolated areas of flower-ing perennial weeds. These are likely to be over-looked when decisions are made about spray ap-plication as they may not be considered important for honeybees particularly if they are feeding on a nearby flowering crop which is more attractive to honeybees. Bumblebees tend to forage on a track-line, i.e. a regular foraging route, and on an individual basis rather than communicating good forage in the way honeybees do. Therefore they tend to visit a relatively small number of flowers on a regular basis. Bumblebees also tend to for-age far closer to the nest than honeybees; 250_300 metres rather than up to 3 km and therefore are more likely to forage on flowering weeds close to their nest.
7.3.	Pesticide incidents
The submission of bumblebees to the Wildlife Incident Investigation Scheme may provide infor-mation on the scale of any problem. Only 3 incidents involving bumblebees have been re-ported in which pesticides were detected, one each in 1995, 1996 and 1997.
1. In 1995 the incident involved dimethoate 0.29 _grbee dimethoate, 0.12 _grbee omethoate.
and may have been linked to an application to oilseed rape in full flower misuse of the pesti-cide. but this could not be confirmed. The bumblebees described as numerous . were found dead and dying in a nearby garden; obviously they are more likely to be detected in a garden than in the field situation.
2. In 1996 0.033 _grbee lambda cyhalothrin was detected in 15 bumblebees after an application to field beans in full flower a misuse of the pesticide .. Although honeybee colonies were situated nearby no deaths occurred. The bum-blebees were detected by an observant bee-keeper who was concerned about the use of the pesticide on a flowering crop.
3. In 1997 alphacypermethrin 0.0044 _grbee. was detected in bumblebees which had been foraging on oilseed rape which had been sprayed whilst in flower. The numbers of bum-blebees found dead was not stated. The spray was applied at 1915 and 1930 i.e. evening . and contained a mixture of alphacypermethrin, car-bendazim and iprodione. Again the bumble-bees were detected two days after the spray application by an observant beekeeper whose apiary was unaffected. Further dead bumble-bees were collected 2 weeks after the the spray application.
Therefore of the three incidents reported in which pesticides were detected two were apparent misuse spray application to a flowering crop. and only one followed normal use. The latter demon-strates the potential for exposure of bumblebees at the time when spraying is recommended as no honeybee colonies were affected.
It is difficult to gauge the scale of the problem from the level of incidents reported. It is unlikely that bumblebee deaths would normally be de-tected unless they are on a large scale, unlikely due to the small size of many colonies. Certainly it is likely that when incidents involving honey-bees are reported deaths of bumblebees are also likely to have occurred but have not been de-tected except in the case of feral colony treat-ment .. However, applications of sprays to flower-ing crops or weeds at times when honeybees are less active are likely to result in unreported bum-blebee deaths. Therefore it is likely that there are
Bumblebee Risk Assessment	163
far more bumblebee deaths than the levels re-ported through the Scheme.
8.	Conclusions
This review has shown that the exposure of bum-blebees to pesticides is likely to be at least that of honeybees. Of particular concern are insecticide applications to flowering crops, such as oilseed rape, at times which, although posing less risk to honeybees, are likely to coincide with foraging bumblebees. In addition, contamination of flower-ing weeds in and around sprayed crops is likely to pose a greater risk to bumblebees than honeybees due to their differing foraging habits and smaller colony size. Therefore, it is important that these differences in foraging behaviour, and thus expo-sure, are considered in the risk assessment pro-cess.
The potential for exposure of the queen bum-blebee early in the season when she is establish-ing her colony is likely to have a greater impact on a bumblebee colony than exposure of workers later in the season, although this is also of con-cern due to the relatively small colony size. Indi-rect effects of pesticides, e.g. herbicides, on bum-blebees are also important through loss of peren-nial weed forage, nesting sites etc.
Given the wide range of plant species depen-dent on bumblebees for pollination impacts on colonies are also likely to have effects on the populations of dependent plant species resulting in less forage. To more closely assess the direct risk posed by pesticides to bumblebees more de-tailed information on their exposure to pesticides applied to crops is required than the hazard ratio currently used for honeybee risk assessment. In addition further information is required to con-firm the limited data suggesting that generally bumblebees are less sensitive to pesticides than honeybees.
Note
1. This does not refer to the threshold used in section 3. There was not enough quantitative information available to use the same threshold for this part of the work, and more qualitative criteria have been used instead.
164 Thompson and Hunt
References
Acker, R.C.V. and Lutman, P.J.W. 1995.. Weed interference in autumn-sown field beans Vicia faba L... In Proceedings of the Brighton Crop Protection Conference_Weeds, 20_23 No¨ember 1995, pp. 907_12. Farnham: BCPC.
Alford, D.V. 1975.. Bumblebees. London: Davis-Poynter Ltd. Banaszak, J. 1984.. The occurrence and number of bees Apodea. on some cultivated crop plants in the Wielkopol-
ska region W. Poland.. Entomologiczne 53, 623_31. Barker, R.J., Lehner, Y. and Kunzmann, M.R. 1980.. Pesti-
cides and honeybees: nectar and pollen contamination in alfalfa treated with dimethoate. Arch. En¨iron. Contam. Toxicol. 9, 125_33.
Barr, C.J., Bunce, R.G.H., Clarke, R.T., Fuller, R.M., Fuse, M.T., Gillespie, M.K., Broom, G.B., Hallam, C.J., Hor-nung, M., Howard, D.C. and Ness, M.J. 1996.. The Coun-tryside Sur¨ey_Main Report. London: Department of the Environment.
Barrow, D.A. 1983.. Ecological Studies on Bumblebees in South Wales with Special Reference to Resource Parti-tioning and Bee Size Variation. PhD Thesis, University College, Cardiff.
Batra, S.W.T. 1993.. Male-fertile potato flowers are selec-tively buzz-pollinated only by Bombus terricola Kirby in upstate New York. J. Kansas Entomol. Soc. 66, 252_4.
Bhalla, O.P., Verma, A.K. and Dhaliwal, H.S. 1983.. Foraging activity of insect pollinators visiting stone fruits. J. Ento-mol. Res. 7, 91_4.
Bishop, J.A. 1994.. Bumblebees Bombus hypnorum. collect aphid honeydew on stone pine Pinus pumila. in the Rus-sian far-east. J. Kansas Entomol. Soc. 67, 220_2.
Bohart, G.E.	1957.. Pollination of alfalfa and red clover.
Annu. Re¨. Entomol. 2, 355_80.
Bohart, G.E. 1960.. Insect pollination of forage legumes. Bee World 41, 51_64.
Brian,A.D. 1957.. Differences in the flowers visited by four species of bumblebees and their causes. J. Animal Ecol. 26, 69_96.
Clarke, J.H., Bowerman, P., Young, J.E.B., Cook, S.K., Jones, A.E., Groves, S.J. and Green, M. 1993.. Effects of recom-mended and reduced rate herbicides on weed number, yield and gross margin in TALISMAN: report on the first two years. In Proceedings of the Brighton Crop Protection Conference_Weeds, 22_25 No¨ember 1993, pp. 1009_14. Farnham: BCPC.
Corbet, S.A., Williams, I.H. and Osborne, J.L. 1991.. Bees and pollination of crops and wild flowers in the European Community. Bee World 72, 47_59.
Corbet, S.A., Fussell, M., Ake, R., Fraser, A., Gunson, C., Savage, A. and Smith, K. 1993.. Temperature and pollinat-ing activity of social bees. Ecol. Entomol. 18, 17_30.
Crane, E. 1990.. Bees and Beekeeping; Science Practice and World Resources. Oxford: Heinemann Newnes.
Davis, A.R. and Shuel, R.W. 1988.. Distribution of carbofu-ran and dimethoate in flowers and their secretion in nectar as related to nectary vascular supply. Can. J. Bot. 66, 1248_55.
Delbrassinne, S. and Rasmont, P. 1988.. Contribution a l’etude de la pollinisation du colza Brassica napus L. var.
oleifera	Moench . Delle, en Belgique. Bull. Rech. Agron.
Gembloux 23, 123_52.
Doorn, A.V. 1993.. Bumblebees breaking through as crop pollinators. In Pollination in Tropics, pp. 195_201. IUSSI.
Dramstad, W. and Fry, G. 1995.. Foraging activity in bumble-bees in relation to flower resources on arable land. Agric. Ecosystems En¨iron. 53, 123_35.
Drescher, W. and Geusen-Pfister, H. 1991.. Comparative testing of the oral toxicity of acephate, dimethoate and methomyl to honeybees, bumblebees and syrphidae. Acta Horticulturae 288, 133_6.
Ellenberg, H. 1988.. Vegetation Ecology of Central Europe, 4th ed. Cambridge: Cambridge University Press.
Emmet, B.J. 1971.. Insect visitors to pear blossom. Plant Pathol. 20, 36_40.
EPPO 1993.. Decision making scheme for the environmental risk assessment of plant protection products_honeybees. Bull. OEPPrEPPO Bull. 23, 151_65.
Firbank, L.G. 1990.. Interactions between weeds and crops. In: Firbank, L.G., Carter, N., Darbyshire, J.F. and Potts, G.R. eds. The Ecology of Temperate Cereal Fields, pp. 469. Oxford: Blackwell Scientific Publications.
Free, J.B. 1968a.. The behaviour of bees visiting runner beans Phaseolus multiflorus.. J. Appl. Ecol. 5, 631_8.
Free, J.B. 1968b.. The foraging behaviour of honeybees Apis mellifera. and bumblebees Bombus spp.. on blackcurrant Ribes nigrum., raspberry Rubus idaeus. and strawberry
Fragaria x ananassa. flowers. J. Appl. Ecol. 5, 157_68. Free, J.B. 1993.. Insect Pollination of Crops, 2nd ed. London:
Academic Press.
Fussell, M., Osborne, J.L. and Corbet, S.A. 1991.. Seasonal and diurnal patterns of insect visitors to winter sown field bean flowers in Cambridge. Aspects Appl. Biol. 27, 95_9.
Fussell, M. and Corbet, S.A. 1991.. Forage for bumblebees and honeybees in farmland; a case study. J. Apic. Res. 30, 87_97.
Fussell, M. and Corbet, S.A. 1992.. Flower usage by bumble-bees: a basis for forage plant management. J. Appl. Ecol. 29, 451_65.
Fussell, M. and Corbet, S.A. 1993.. Bumblebee Hym. Api-dae. forage plants in the United Kingdom. Entomologists Monthly Magazine 129, 1_14.
Glasgow, J.L., Dicks, J.W. and Hodgson, D.R. 1976.. Compe-tition by, and chemical control of, natural weed popula-tions in spring-sown field beans Vicia faba.. Ann. Appl. Biol. 84, 259_69.
Goodwin, S.G. 1995.. Seasonal phenology and abundance of early-, mid- and late-season bumblebees in Southern Eng-land 1985_1989. J. Apic. Res. 34, 79_89.
Greig-Smith, P.W., Thompson, H.M., Hardy, A.R., Bew, M.H., Findlay, E. and Stevenson, J.H. 1994.. Incidents of poison-ing of honeybees Apis mellifera. by agricultural pesticides in Great Britain 1981_1991. Crop Protection 13, 567_81.
Gretenkord, C. and Drescher, W. 1993.. Effects of four pesticides Decis, Metasystox, Pirimor, Rubitox. on the bumblebee Bombus terrestris L.: determination of the oral LD50 and preliminary results with semi-field tests. Api-dologie 24, 519_20.
Griffiths, D. and Robberts, E.J. 1996.. Bumblebees as polli-nators of glasshouse crops. In Matheson, A. ed.. Bumble ees for Pleasure and Profit, pp. 33_9. Cardiff: IBRA.
Gubin, A.F. 1945.. Cross-pollination of fibre flax. Bee World 26, 30_1.
Gurr, L. 1975.. The role of bumblebees as pollinators of red clover and lucerne in New Zealand; a review and prospect. Proc. N.Z. Grassland Association 36, 111_22.
Gwynne, D.C. and Murray, R.B. 1985.. Weed Biology and Control in Agriculture and Horticulture. London: Batsford Technical.
Gyan, K.Y. and Woodell, S.R.J. 1987.. Analysis of insect pollen loads and pollination efficiency of some common insect visitors of woody Rosaceae. Functional Ecol. 1, 269_74.
Harder, L.D. 1983.. Functional differences of the proboscides of the short- and long-tongued bees Hymenoptera, Apoidae.. Can. J. Zool. 61, 1580_6.
Heinrich, B. 1979.. ‘‘Majoring’’ and ‘‘minoring’’ by foraging bumblebees, Bombus ¨agans; an experimental analysis. Ecology 60, 245_55.
Heinrich,	B.	1993 ..	The	Hot-Blooded	Insects,	Berlin:
Springer-Verlag.
Herrera, C. 1990.. Daily patterns of pollinator activity, differ-ential pollinating effectiveness and floral resource avail-ability in a summer flowering Mediterranean shrub. Oikos 58, 277_88.
Holm, N. 1966.. The utilization and management of bumble-bees for red clover and alfalfa seed production. Annu. Re¨. Entomol. 11, 155_77.
Hurd, P.D. 1964.. The pollination of pumpkins, gourds and squashes genus Curcubita.. Bee World 47, 97_8.
Inglesfield, C. 1989.. Pyrethroids and terrestrial non-target organisms. Pestic. Sci. 27, 387_428.
Jaycox, E.R. 1964.. Effect on honeybees of nectar from systemic insecticide-treated plants. J. Econ. Entomol. 57, 31_5.
Kendall, D.A. and Smith, B.D. 1976.. The pollinating effi-ciency of honeybee and bumblebee visits to flowers of the runner bean Phaseolus coccineus L. J. Appl. Ecol. 13, 749_52.
Kendall, D.A. and Solomon, M.E. 1973.. Quanitites of pollen on the bodies of insects visiting apple blossom. J. Appl. Ecol. 10, 627_34.
Kevan, P.G., Straver, W.A., Offer, M. and Laverty, T.M. 1993.. Pollination of greenhouse tomatoes by bumblebees in Ontario, Canada. Proc. Entomol. Soc. Ontario 3, 4_6.
Knott, C.M., May, M.J. and Ward, J.T. 1995.. Weed control in potatoes, oilseed rape, pulses and sugar beet_trends and prospects. In Proceedings of the Brighton Crop Protec-tion Conference_Weeds, 20_23 No¨ember 1995, pp. 1193_202. Farnham: BCPC.
Koch, H. and Weiber, P. 1997.. Exposure of honeybees during pesticide application under field conditions. Api-dologie 28, 439_47.
Legeun, J., Mesquida, J., Pierre, J.S., Morin, G., Tasei, J.N. and Carre, S. 1993.. Efficiency of pollinating treatments on 2 lines of spring faba bean Vicia faba L. var equina Steudel. with different levels of self fertility and using
Bumblebee Risk Assessment	165
three species of	Bombus Latr.	Hymenoptera, Apidae..
Apidologie 24, 129_45.
Longley, M. and Sotherton, N.W. 1997.. Factors determining the effects of pesticides upon butterflies inhabiting arable farmland. Agric. Ecosystems En¨iron. 61, 1_12.
Lord, K.A., May, M.A. and Stevenson, J.H. 1968.. The secre-tion of the systemic insecticides dimethoate and phorate into nectar. Ann. Appl. Biol. 61, 19_27.
Lundberg, H. 1980.. Effects of weather on foraging flights of bumblebees Hymenoptera, Apidae. in subalpineralpine area. Holarctic Ecol. 3, 104_10.
Lutman, P.J.W., Bowerman, P., Palmer, G.M. and Whytock, G.P. 1995.. A comparison of the competitive effects of eleven weed species on the growth and yield of winter oilseed rape. In Proceedings of the Brighton Crop Protection Conference_Weeds, 20_23 No¨ember 1995, pp. 877_82. Farnham: BCPC.
Macdonald, M.A. 1998.. The feeding ecology of some Bom-bus and Psithyrus bees in North Scotland and the impor-tance to them of native and exotic plants. In preparation.
Macfarlane, R.P., Griffin, R.P. and Read, P.E.C. 1983.. Bum-blebeemanagement options to improve ‘‘Grasslands Paw-era’’ red clover seed yields. Proc. N.Z. Grassland Associa-tion 44, 47_53.
Macfarlane, R.P., Ende, H.J. and Griffin, R.P. 1991.. Pollina-tion needs of ‘‘Grasslands Pawera’’ red clover. Acta Horti-culturae 288, 399_404.
MAFF 1997.. http:rrwww.maff. go¨.uk General Agricultural Statistics Section..
Marshall, E.J.P. 1989.. Distribution patterns of plants associ-ated with arable field edges. J. Appl. Ecol. 26, 247_57.
Mayer, D.F. 1983.. Behaviour of pollinators on Malus. In Proceedings of the 5th International Pollination Symposium, Versailles, pp. 387_390.
McGregor, S.E. 1976.. Insect Pollination of Culti¨ated Crop Plants. Washington DC: Agricultural Research Service, USDA.
Michaelson-Yeates, T.P.T., Marshall, A.H., Williams, I.H., Carreck, N.L. and Simpkins, J.R. 1997.. The use of isoen-zyme markers to determine pollen flow and seed paternity mediated by Apis mellifera and Bombus spp. in Trifolium repens, a self-compatible plant species. J. Apic. Res. 36, 57_62.
Mortimer, A.M. 1990.. The biology of weeds. In Hance, R.J. and Holly, K. eds. Weed Control Handbook: Principles, 8th ed, pp. 582. Farnham: BCPCrBlackwell.
Mountford, J., Parish, T. and Sparks, T.H. 1994.. The flora of field margins in relation to land use and boundary fea-tures. In Boatman, N. ed. Field Margins; Integrating Agri-culture and Conser¨ation, Monograph 58, pp. 105_10. Farn-ham: BCPC.
Nye, W.P. and Andersen, J.L. 1974.. Insect pollinators fre-quenting strawberry blossoms and effect of honeybees on yield and fruit quality. J. Amer. Soc. Hortic. Sci. 99, 40_4.
Olgivy, S.E., Green, M.R., Groves, S.J. and Jones, A.E. 1993.. SCARAB_the impact of less intensive herbicide use on the diversity and distribution of weed species in three arable rotations. In Proceedings of the Brighton Crop Protec-tion Conference_Weeds, 22_25 No¨ember 1993, pp. 1241_6. Farnham: BCPC.
166 Thompson and Hunt
Osborne, J.L. 1994.. Evaluating a Pollination System: Borago officinalis and bees. PhD Thesis, University of Cambridge.
Pallutt, B. 1993.. Population dynamics and competition of weeds depending on crop rotation and mechanical and chemical control measures in cereals. In Proceedings of the Brighton Crop Protection Conference_Weeds, 22_25 No¨ember 1993, pp. 1197_204. Farnham: BCPC.
Patten, K.D., Shanks, C.H. and Mayer, D.F. 1993.. Evaluation of herbaceous plants for attractiveness to bees for use near cranberry farms. J. Apic. Res. 32, 73_9.
Pinchinat, B., Bilinski, M. and Ruszkowski, A. 1979.. Possibil-ities of applying bumblebees as pollen vectors in tomato F1 hybrid seed production. In Proceedings of the 4th Inter-national Symposium on Pollination, Maryland, pp. 73_90.
Plowright, R.C. and Laverty, T.M. 1984.. The ecology and sociobiology of bumblebees. Annu. Re¨. Entomol. 29, 175_99.
Pojar, J. 1973.. Pollination of typically anemophilous salt marsh plants by bumblebees, Bombus terricola occidentalis Grne. Amer. Midland Naturalist 89, 448_51.
Proctor, M., Yeo, P. and Lack, A. 1996.. The Natural History of Pollination. London: Harper Collins.
Prys-Jones, O.E. 1982.. Ecological Studies of Foraging and Life History in Bumblebees. PhD Thesis, Cambridge Uni-versity.
Prys-Jones, O.E. and Corbet, S.A. 1991.. Bumblebees, 2nd ed.
Slough: Richmond Publishing.
Reader, R.J. 1977.. Bog ericad flowers: self compatibility and attractiveness to bees. Can. J. Botany 55, 2279_87.
Richards, K.W. and Edwards, P.D. 1988.. Density, diversity and efficiency of pollinators of sainfoin, Onobrychis ¨iciifo-lia sp. Can. Entomoloist 120, 1085_100.
Sanford, J.C. and Hannemann, R.E. 1981.. The use of bees for the purpose of inter-mating in potato. Amer Potato J. 58, 481_5.
Saville, N.M. 1993.. Bumble Bee Ecology in Woodlands and Arable Farmland. PhD Thesis, Cambridge University.
Seeley, T.D. 1985.. Honeybee Ecology, Princeton NJ: Prince-ton University Press. Shipp, J.L., Whitfield, G.H. and Pa-padopoulos, A.P. 1994.. Effectiveness of the bumblebee, Bombus impatiens Cr. Hymenoptera, Apidae. as a pollina-tor of the greenhouse sweet pepper. Scientia Horticulturae 57, 29_39.
Stace, C.A.	1997..	New Flora of the British Isles, 2nd ed.
Cambridge: Cambridge University Press.
Stevenson, J.H. and Racey, P.A. 1966.. Toxicity of Insecti-cides to Bumblebees. Report Rothamsted Experimental Station 176.
Stoddard, F.L. and Bond, D.A. 1987.. The pollination re-quirements of the faba bean. Bee World 68, 144_52.
Stone, G.N. and Willmer, P.G. 1989.. Warm-up rates and body temperature in bees; the importance of body-size, thermal regime and phylogeny. J. Exp. Biol. 147, 303_28.
Tasei, J.N., Sabik, H. and Pirastru, L. 1994.. Effects of sublethal doses of deltamethrin Decis ce . on Bombus terrestris. J. Apic. Res. 33, 129_35.
Teras, I. 1976.. Bumblebees, Bombus Latr. Hymenoptera, Apidae. on red clover in South Savo, Finland. Ann. Agric. Fenniae 15, 116_27.
Teras, I.	1985.. Food plants and flower visits of bumblebees
Bombus: Hymenoptera, Apidea. in Southern Finland. Acta Zool. Fennica 179, 1_120.
van der Steen, J.J.M. 1994.. Method development for the determination of the contact LD50 of pesticides to bum-blebees Bombus terrestris L... Apidologie 25, 463_5.
Voloshin, M., Voloshina, T. and Valter, V. 1995.. Use of wild bees for alfalfa pollination. In Banaszak, J. ed. Changes in Fauna of Wild Bees in Europe, pp. 220. Bydgoszez: Peda-gogical University.
Walton, C.L.	1927.. Note on the activities of bumblebees
Bombus. in North Wales. Ann. Appl. Biol. 14, 465_9.
Watkinson, A.R. and Bo, L. 1993.. Patterns of abundance in the weed seed bank. In Proceedings of the Brighton Crop Protection Conference_Weeds, 22_25 No¨ember 1993, pp. 293_298. Farnham: BCPC.
Williams, C.S. 1997.. Foraging Ecology of Nectar-Collecting Bumblebees and Honeybees. PhD Thesis, University of Cambridge.
Williams, I.H. 1985.. The pollination of swede rape Brassica napus L... Bee World 66, 16_22.
Williams, I.H. 1987.. The pollination of lupins. Bee World 68, 10_6.
Williams, I.H. 1988.. The pollination of linseed and flax. Bee World 69, 145_52.
Williams, I.H., Martin, A.P., Ferguson, A.W. and Clark, S.J. 1990.. Effect of pollination on flower, pod and seed pro-duction in the white lupin Lupinus albus.. J. Agric. Sci. 115, 67_73.
Williams, I.H., Carreck, N. and Little, D.J. 1993.. Nectar sources for honeybees and the movement of honeybee colonies for crop pollination and honey production in England. Bee World 74, 160_75.
Williams, I.H. and Free, J.B. 1974.. The pollination of onion Allium cepa L.. to produce hybrid seed. J. Appl. Ecol. 11, 409_18.
Willmer, P.G. 1983.. Thermal constraints on activity patterns in nectar feeding insects. Ecol. Entomol. 8, 455_69.
Willmer, P.G., Bataw, A.A.M. and Hughes, J.P. 1994.. The superiority of bumblebees to honeybees as pollinators; insect visits to raspberry flowers. Ecol. Entomol. 19, 271_84.
Wilson, G.F. 1929.. Pollination of hardy fruits: insect visitors to fruit blossoms. Ann. Appl. Biol. 16, 602_29.
Yalden, P.E. 1982.. Pollen collected by the bumblebee Bom-bus monticola Smith. in the peak district, England. J. Natural History 16, 823_32.